powertop suggests to remove g-p-m

Bug #117745 reported by Nicolò Chieffo
28
This bug affects 2 people
Affects Status Importance Assigned to Milestone
gnome-power-manager (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

Binary package hint: gnome-power-manager

the new intel utility powertop suggests to remove gnome-power-manager:

Suggestion: Disable or remove 'gnome-power-manager' from your system.
Despite its name, some versions of gnome-power-manager end up costing more power
than it'll ever save.

The website http://www.linuxpowertop.org/known.php also suggests to remove gnome-screensaver

Revision history for this message
jcfp (jcfp) wrote :

Thank you for your report. I don't see any package for powertop in the ubuntu repositories. How exactly does this suggestion by the powertop developers constitute a bug in the gnome-power-manager package?

Changed in gnome-power-manager:
status: Unconfirmed → Needs Info
Revision history for this message
Nicolò Chieffo (yelo3) wrote : Re: [Bug 117745] Re: powertop suggests to remove g-p-m

I should have written that I'm using gutsy.
If you don't know what powertop is, it is a new intel utility that
discovers which process drains the battery life

this is the site: http://www.linuxpowertop.org

Revision history for this message
Thomas Leonard (talex5) wrote :

Stracing it, it looks like the main culprit is this (in gnome-power-manager-2.18.2/src/gpm-dpms.c):

static void
gpm_dpms_init (GpmDpms *dpms)
{
 dpms->priv = GPM_DPMS_GET_PRIVATE (dpms);

 add_poll_timer (dpms, 500);
}

This wakes up the CPU twice a second to check the monitor's DPMS state.

Revision history for this message
Richard Hughes (richard-hughes) wrote :

Just update to 2.18.3, I've made the poll less aggressive.

Revision history for this message
Nicolò Chieffo (yelo3) wrote :

I think that there should be a communication between powertop
developers and gnome-power-manager developers... It is useless to make
a software to save power if it does not save power...

Revision history for this message
Richard Hughes (richard-hughes) wrote :

>It is useless to make a software to save power if it does not save power

It does save power. The real problem was that DPMS does not give async events on chance, therefor we had to poll. Now the X guys are using xrandr for per-head DPMS we can do this without polling at all. Update g-p-m and watch the recommendation go away.

Revision history for this message
Nicolò Chieffo (yelo3) wrote :

ok thanks for the explanation

Changed in gnome-power-manager:
status: Incomplete → Fix Released
Revision history for this message
Vadim Peretokin (vperetokin) wrote :

While g-p-m was fixed, powertop still gives the recommendation to kill it once in a while.

And if you do kill it, then the suspend/resume/reboot menu's go away.

Revision history for this message
Paul Rensing (paulrensing) wrote :

I see the same thing (Ubuntu 8.10, powertop 1.10-1, gnome-power-manager 2.24.0-0ubuntu8.1). You only see it if you follow up on the other suggestions powertop makes first.

You can definitely see the issue. In powertop's stats, gnome-power-manager is responsible for ~20% of the wake-ups, second only to my wireless card. This is with a fairly quiet system.

Revision history for this message
Flavio Tordini (zarlino) wrote :

Yes, that still happens in intrepid and actually powertop sometimes reports that gnome-power-manager does many wakeups.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.